ML24071A123

Staff’S Assessment of Hazards from a Hypothetical Rupture and Subsequent Release of Natural Gas of the Three Rivers Pipeline Lateral to the Dresden Power Station Units 2 and 3

ML24071A123
Person / Time
Site:	Dresden
Issue date:	03/07/2024
From:	NRC/RGN-III/DORS/RPB1
To:
References
Download: ML24071A123 (1)

Text

1 Staffs Assessment of Hazards from a Hypothetical Rupture and Subsequent Release of Natural Gas of the Three Rivers Pipeline Lateral to the Dresden Power Station Units 2 and 3

1.0 INTRODUCTION

The Three Rivers Lateral (TRL) pipeline is part of the proposed IL-19 Three Rivers Interconnection project in Illinois of the Alliance Pipeline L.P. This pipeline brings natural gas from a point east of the Aux Sable Compressor Station in the north to a new gas-fired power generation facility in Goose Lake Township in the south. Dresden Nuclear Power Station (Dresden Station) has assessed the potential hazards this TRL pipeline could pose to the Dresden Units 2 and 3 and documented the analyses and results in a report, Dresden Buried Pipe Blast and Gas Leak Analysis. Calculation No. DRE22-0002. Revision 001. September 7, 2023. This report discusses the expected annual failure frequency of the TRL pipeline and has estimated the hazards from the overpressure and the thermal radiation at three locations of interest of the Dresden Station from a hypothetical rupture of the pipeline. The NRC staff has reviewed the safety assessments presented in this report. The staff review and findings are given below.

2.0 BRIEF DESCRIPTION OF THE FACILITY AND THE PIPELINE The proposed TRL pipeline and the nearby Dresden Station are described below.

2.1 Description of the Facility This pipeline will traverse roughly north to south through west of the Independent Spent Fuel Storage Installation (ISFSI) operated by the GE Hitachi Nuclear Energy as the Morris Operation, as shown in Figure 1 of Dresden (2023). This ISFSI is west of both Dresden Unit 2 and Unit 3.

Therefore, both units of Dresden Station will be further away from the pipeline compared to the GE Hitachi Nuclear Energy Morris Operation.

The reactor building at the Dresden Station encloses the reactor, the primary containment, and most of the equipment associated with safe operation of the reactor. The reactor building has a steel-reinforced concrete bottom portion with a structural steel superstructure. The reactor building is designed for protection against tornadoes and tornado generated missiles and can withstand a 6.3-psi [43.4-kPa] pressure drop during a tornado event (Dresden UFSAR, 2009).

Dresden (2023) has assumed that the rupture takes place at the worst location of the pipeline, at a point closest to the Dresden Station. Consequently, the estimated hazards (overpressure, heat flux, and methane concentration) would be for the worst-case scenario. The shortest distances of the TRL pipeline from three locations of interest within the Dresden Station are given in Figures 2 through 4 of Dresden (2023). The edge of the nearest plant structure is 3,366 ft or 0.64 mi [1.03 km] away and the Control Room intake is 3,701 ft or 0.70 mi [1.13 km]

away from the proposed TRL pipeline route. The Meteorological Tower is closest to the pipeline and is 1,259 ft or 0.24 mi [0.38 km] away.

2 2.2 Description of the Pipeline The length of the TRL pipeline is 2.85 mi [4.6 km] with an outer diameter of 20 in. [50.8 cm] and a wall thickness of 0.375 in. [0.953 cm. The maximum design operating pressure of the pipeline is 1,208 psig [8,330 kPa]. The natural gas will be at temperatures between 70° and 90 °F [21° and 32° C]. The pipeline will have an internal surface roughness of 300 µin./in. (7.62 µm/m), as stated in Table 3 of Dresden (2023).

The proposed TRL pipeline will have NPS 20 automated isolation valves installed at each end of the pipeline. It would take approximately 20 s to close these automated valves. As stated in Table 3 of Dresden (2022), a delay of at least 10 minutes has been assumed before the ruptured section would be isolated by closure of the isolation valves, following Factory Mutual Datasheet 7-42 (Factory Mutual Insurance Company, 2008).

2.3 Natural Gas in the Pipeline The natural gas to be transmitted through this TRL pipeline will be mostly methane, 90.69 percent (%) by volume. Rest of the components are ethane 6.6%, propane 0.47%, and other hydrocarbons, as given in Table 3 of Dresden (2023). In addition, small amounts of nitrogen and carbon dioxide will be present. The molecular weight of the gas mixture is 17.45 lb/lb-mole [17.45 kg/kg-mol] and the net heat of combustion is 20,630 Btu/lb

[47,990 kJ/kg], as stated in Table 3 of Dresden (2023). This natural gas is slightly richer (less methane) with a lower heat of combustion than the pure methane 21,509 Btu/lb [50,030 kJ/kg]

(Table 15-2, NRC, 2004). Dresden (2003) has assumed that this natural gas is pure methane (100% methane). The NRC staff accepts this assumption as it will increase the heat of combustion in the analysis resulting in slightly higher estimated heat flux at any location.

3.0 POTENTIAL CONSEQUENCES OF THE PIPELINE RUPTURE A rupture of a natural gas transmission pipeline is an extremely rare event; more likely scenario would be a leak from small openings in the pipeline. Nevertheless, Dresden (2023) has conducted an analysis to assess the potential consequences of a full line rupture (i.e., a guillotine break of the pipeline) resulting in natural gas escaping simultaneously from both ends of the severed pipeline. This is the worst-case release scenario from a pipeline and bounds the consequences of leaks.

A rupture of a high-pressure natural gas pipeline rapidly releases the gas through the break of the pipeline in the form of a jet. In case of a buried pipeline, the rupture of the pipeline may eject the overburden and form a crater. That is why a jet fire from a buried pipeline is sometime called a crater fire. The consequences from a guillotine break of a natural gas pipeline can be divided into (Sluder et al., 2022):

Missile generation Flash fire, in case of delayed ignition Jet fire, in case of immediate ignition Overpressure from explosion.

3 Dresden (2023) did not consider the potential consequences from a missile generated from a hypothetical rupture of the TRL pipeline. The staff reviewed reported distance traveled by pipeline fragments, as tabulated in Table 1 of NRC (2020), and pipeline accident reports of the National Transportation Safety Board (NTSB). The staff finds that the pipeline fragment travel distance rarely exceeds the Potential Impact Radius (PIR), as defined in 49 CFR Part 192.903.

The travel distance of the pipeline fragments in most cases is less than 50% of the PIR. The staff notes that the travel distance in rare cases exceeds the PIR distance, but the exceedance distance is less than 10% of the PIR. The PIR for the TRL pipeline 508 ft or 0.096 mi [0.155 km]

considering the gas composition is close to the Rich Natural Gas, as given in Baker (2005).

Assuming the released gas is Lean Natural Gas (pure methane), the PIR is 480 ft or 0.091 mi

[0.146 km].

The staff finds that the distance to all three locations of interest at the Dresden Station significantly exceeds the PIR distance and also the maximum recorded travel distance of fragments generated in a pipeline rupture. In addition, the staff notes that the safety-related structures at the Dresden Station are designed against a strike by a utility pole as a tornado missile. The analyzed missile weighs 1,490 lb [676 kg] and strikes a structure at a speed of 144 mph [232 km/h] (Dresden UFSAR, 2009). The momentum of such a missile far exceeds the missiles generated in a pipeline rupture, especially at the instant of striking a safety-related structure located at a significant distance away from the rupture location. Based on the preceding discussion, the staff concludes that missiles generated from rupture of the TRL pipeline are not expected to damage the structures and be a credible hazard to the Dresden Station.

A flash fire results from relatively slow (less than the speed of sound) burning (or, deflagration) of the released natural gas plume in open air. Short duration of the flame produced by deflagration of the released gas precludes serious damage to the safety-related concrete structures. A flash fire may burn back to the source (the ruptured section of the pipeline) becoming a jet fire. Sluder et al. (2022) states that 74% of ignition of flash fires occurs within 2 min of the pipeline rupture and the fire typically lasts only for a short time, 30 s of less. The overpressure generated from a flash fire is insignificant because of deflagration in an unconfined (open) space, rather than detonation, of the natural gas. Therefore, the staff finds it acceptable to eliminate consequences of the flash fire after rupture of the proposed TRL pipeline from further consideration.

A jet fire can have more severe consequences (CCPS, 2014) and results from rapid release of the natural gas through the ruptured area of the pipeline in the form of a momentum jet.

Immediate ignition of the released natural gas-air mixture results in a jet fire. Sparks generated by the fragments of the ruptured pipe and/or rock particles ejected can potentially ignite the released natural gas and starts the jet fire. Within a very short time after ignition, the jet fire reaches its full intensity. The heat flux decreases as soon as the pipeline isolation valves are closed as the flow of natural gas declines. Fire will continue until the gas released from the pipeline is consumed.

4 A vapor cloud explosion (VCE) occurs when a cloud of flammable gas ignites, and the flame speed accelerates to sufficiently high velocities to develop significant overpressure. The conditions necessary for a VCE of the released natural gas to occur are (Sluder et al., 2022):

1. Ignition must be delayed so that the natural gas-air cloud can form an ignitable mixture with concentration within the flammable range of natural gas, i.e., between 5% (the lower flammability limit or LFL) and 15% (the upper flammability limit or UFL).

2. Availability of confinement with significant congestion in the flame propagation path where the flammable natural gas is released.

3. Ignition source must have sufficient energy to ignite the natural gas-air mixture.

Release of natural gas from a transmission pipeline, if ignited, typically results in a rapid burning fire or deflagration, rather than a detonation or explosion. In a deflagration, the flame propagates through the unburned natural gas-air mixture at a speed slower than the speed of sound, as stated before. A detonation event, where the flame propagates at supersonic velocity, can generate significant overpressure. Overpressure developed in a deflagration event is dependent on the rate of combustion of the fuel-air mixture and other factors (Tang and Baker, 1999), and lower than that in a similar detonation event. A detonation may result directly if the released natural cloud is ignited with sufficient energy (direct detonation). Alternatively, a detonation can take place if the propagating flame of the released natural gas cloud accelerates to the speed of sound under certain conditions, such as availability of adequate confinement and congestion (obstacles especially of repeated nature in the flame propagation path generating turbulence and increasing flame speed without preventing expansion) resulting in deflagration to detonation transition (DDT).

The NRC staff notes that the availability of an ignition source with sufficient energy to initiate a direct detonation of the released natural gas cloud is extremely rare. Ignition energy needed to deflagrate methane-air mixture is of the order of 107 Btu [104 J]; however, direct initiation of detonation requires approximately 105 Btu [108 J] of energy, an increase of 12 orders of magnitude (Mercx and van den Berg. 2005). Such amount of energy may be available from high explosives. However, common ignition sources, such as sparks from ejected pipeline fragments or electrical apparatus, hot steam lines, open furnaces, heaters, and moving parts in a machinery, do not generally possess the amount of energy to directly initiate a detonation event.

Therefore, direct initiation of detonation of natural gas vapor cloud even near the release point is an extremely unlikely phenomenon.

If ignition of the vapor cloud occurs immediately after the pipeline rupture, a jet fire would make detonation of the released natural gas highly unlikely. If the ignition of the plume occurs late, a flash fire may occur which may revert back to the ruptured location to become a jet or crater fire.

Additionally, the natural gas plume is buoyant because methane has lower molecular weight than either nitrogen or oxygen in air. The buoyant nature of the plume generally precludes a persistent flammable plume of natural gas at the ground level. As the plume rises, availability of structures capable of providing sufficient confinement becomes less likely. Because of the buoyancy, the plume may be above the structures when it is near the Dresden Station reducing

5 the likelihood of finding confinement for natural gas-air mixture. In addition, sufficient congestion may not be available, an important precondition for initiating a detonation. Additionally, methane being a low reactivity fuel (Mercx and van den Berg, 2005), has low propensity to flame acceleration under confinement and congestion, and, as a consequence, generates quite low overpressure. If the vapor cloud is not ignited, the methane concentration in the plume rapidly dilutes to the LFL limit and the buoyant plume dissipates into the surrounding atmosphere.

Figures 1 through 4 of Dresden (2023) show that the area between the route of the TRL pipeline and the Dresden Station is open space lacking any structure capable of confining the released natural gas in sufficient quantities. In addition, there is no congestion in the propagation path of the gas cloud to reach the Dresden Station. Therefore, the staff notes that due to lack of sufficient confinement and congestion from the potential rupture point and the Dresden Station, any potential transition of a deflagration event to a detonation would be extremely unlikely and the generated overpressure from deflagration would be very low.

A review of major vapor cloud incidents around the world (Atkins et al., 2017), commissioned by the Pipeline and Hazardous Materials Safety Administration (PHMSA) of the Department of Transportation (DOT) and the UK Health and Safety Executive (HSE), did not identify any historical records of VCEs of liquefied natural gas or methane in open areas with sufficient severity to cause damage. The NRC Expert Evaluation Team (NRC, 2020) contacted the PHMSA pipeline accident investigators in connection with potential rupture and associated consequences of natural gas pipelines near the Indian Point Energy Center nuclear powerplants (NRC, 2020). The NRC Expert Evaluation Team noted that these accident investigators had expressed a similar opinion to that of the aforementioned worldwide study (Atkinson, et al.,

2017) in that they were unaware of any large natural gas (methane) delayed vapor cloud explosions from rupture of a pipeline. In addition, the NRC Expert Evaluation Team did not find any record of dense methane gas clouds igniting or exploding at a location away from the initial pipe rupture.

Based on the preceding discussion, the staff finds that assessment of the potential consequences of only jet fire and overpressure events after a hypothetical rupture of the proposed TRL pipeline is appropriate. A flash fire or the potential missiles generated from a pipeline rupture are not expected to cause significant damage to the safety-related structures at the Dresden Station because of the large intervening distance and the design of these structures.

4.0 ANALYSIS OF THE HAZARDS As discussed before, potential hazards from a hypothetical rupture of the TRL pipeline include jet/crater fire emitting radiative heat flux if the ignition of the released natural gas vapor cloud is near-instantaneous and overpressure if the ignition is delayed. Dresden (2023) has estimated the annual frequencies of rupture of the TRL pipeline. Dresden (2023) has also estimated the maximum overpressure and the sustained thermal flux to be experienced at the three pre-selected locations of the Dresden Station. In addition, Dresden (2023) has estimated the methane concentration outside and inside the Control Room of the Dresden Station. The NRC

6 staff has reviewed these analyses. The staff review along with the findings are summarized below.

4.1 Annual Frequency of Pipeline Rupture Dresden (2023) has reported the annual frequency of failure of the natural gas transmission pipeline to be 7.8 x 105 failures per mile for failure sizes varying from 2 in. [5.1 cm] to full bore rupture for pipelines having diameter from 14 in. [35.6 cm] to 20 in. [51 cm], based on information from J. House Environmental (2004). Dresden (2023) has assumed that the probability a rupture of the TRL pipeline at or close to the point nearest to the Dresden Station would be very low because of small frequency of rupture of the pipeline, approximately 8 x 105 failures/mi/yr.

Staff Assessment The staff has reviewed the information presented in J. House Environmental (2004) and finds that the information of pipe rupture frequency is based on data from early 1980s and even earlier from the Department of Transportation (DOT). The highest failure rate in each natural gas transmission pipeline diameter range occurs for hole size less than 0.25 in. [0.64 cm] in diameter, a leak rather a guillotine break of the pipeline. Failures with larger hole sizes occur less frequently with the full rupture of the pipelines being the rarest events. The staff has also reviewed the information and discussion given in Appendixes C and D of NRC (2020) and finds that for natural gas transmission lines with diameter 20 in. [51 cm] or more and maximum operating pressure of 300 psi [2,068 kPa] or more, the frequency of pipeline failure is estimated to be 2.4 x 105 failures per mile per year. This estimation used more recent (2002 through 2019) pipeline failure data from the DOTs Pipeline and Hazardous Materials Safety Administration (PHMSA). Therefore, the staff finds that the annual frequency of full rupture of the pipeline assumed by Dresden (2023) is conservative.

4.2 Overpressure Estimation A delayed ignition of the flammable region of the released natural gas vapor cloud can generate overpressure if all the conditions necessary for a VCE, as discussed before, are fulfilled.

Dresden (2023) has estimated the overpressure from a potential VCE using the ALOHA computer program, Version 5.4.4, developed by the Environmental Protection Agency (EPA) and the National Oceanic and Atmospheric Administration (NOAA). The computer program is under Enercons Quality Assurance program. The staff has reviewed the analysis presented and assessment of consequences, as discussed below. In addition, the staff has conducted a confirmatory analysis using analytical solutions available in the literature.

4.2.1 Methodology Used An ignition of the natural gas plume within the flammable limits (between 5% and 15%) can potentially generate an explosion. As discussed before, direct detonation of the released natural gas is not possible as natural gas (methane) requires about 12 orders of more energy to detonate than to deflagrate (rapid burning). Energy from commonly available ignition sources has too low energy to detonate a natural gas cloud. As the natural gas is a low reactive material, DDT is also not a possibility because the flame speed developed from highest confinement

7 (2-D) (released gas is confined between two parallel rigid surfaces, which is not there at the site) and high congestion (obstacles blocking more than 40% area in the flame propagation path, a scenario not applicable at the site) would not be sufficient to accelerate the flame speed from deflagration to detonation, as shown in Table 2 of Pierorazio et al. (2005).

Trinitrotoluene (TNT) is a high energy-density explosive that undergoes detonation in which the reaction front travels at sonic velocity or higher. In contrast, the natural gas cloud is a low energy-density source that undergoes deflagration in which the flame front propagates at a much slower velocity than the sonic velocity. As discussed before, a wide spectrum of flame speed in the hydrocarbon vapor cloud can result from flame acceleration under a variety of confinement and congestion scenarios in the flame propagation path but the flame speed is never observed to reach the sonic velocity. The TNT Equivalent method inherently assumes the flame propagates at the sonic velocity and converts the energy available to an equivalent amount of TNT and estimates the overpressure generated from detonation. A major drawback of the TNT Equivalent method is that the estimated overpressure is directly proportional only to the amount of fuel involved and completely ignores the effects of combustion mode of the hydrocarbon vapor. In other words, the TNT Equivalent method assumes completely different phenomena and is not appropriate for use with hydrocarbon vapor cloud explosion based on overwhelming evidence worldwide (Atkinson, 2017; NRC, 2020).

The ALOHA computer program uses the Baker-Strehlow-Tang (BST) blast curve-based method, instead of the TNT Equivalent method, to estimate the overpressure from delayed ignition of the natural gas vapor cloud generated following a hypothetical guillotine rupture of a pipeline. The BST method predicts the overpressure after ignition from either a low speed (subsonic) burn (deflagration) or a high-speed (sonic) burn (detonation) of the flammable region of the natural gas cloud. The maximum positive overpressure generated within the flammable region of the vapor cloud is independent of the fuel mass within this region. Instead, the overpressure depends on the flame propagation speed within the cloud, which, in turn, depends on the reactivity of the material (here, the natural gas), degree of confinement available to the flammable portion of the cloud, and congestion or degree of obstacles in the flame propagation path. Outside the flammable region of the vapor cloud, the overpressure depends on the total fuel mass within the flammable region of the cloud and the flame propagation speed. The overpressure decreases as the distance increases outside the cloud.

4.2.2 Scenarios Modeled Dresden (2023) has assumed a wind speed of 2.31 mph [3.7 kph] blowing from the southwest toward the Dresden Station with the Pasquill atmospheric stability class F for the analysis. This means the rupture takes place at night with overcast conditions (50% cloud cover assumed).

The assumed wind speed is the lowest wind speed the ALOHA program will take as input. The actual wind speed at the site exceeds this speed 84% of the time (Dresden, 2023). Additionally, the wind blows only 6.6% of time from the southwest direction at the Dresden Station. A nighttime temperature of 75 °F [24 °C] in August was selected in the analysis. In addition, the ground between the pipeline and the Dresden Station is taken as Open Country, as shown in Figure 1 of Dresden (2023).

8 The staff finds that the low wind speed with the atmospheric stability F and wind blowing from the southwest is extremely rare at the Dresden Station, only happens 0.39% of time (Dresden, 2023). The staff also finds that Pasquill stability class F generates a condition not favorable for easy dispersion of the vapor cloud of natural gas generated after the rupture of the pipeline.

Absence of solar insolation and Open Country will also reduce dispersion of the released natural gas in the atmosphere and produce a worst-case condition for overpressure and thermal radiation. Based on the preceding discussion, the staff concludes that Dresden (2023) has selected the atmospheric conditions that would result in worst-case conditions for generating overpressure, methane concentration, and thermal radiation although the assumed scenarios would occur extremely rarely.

A guillotine rupture of the pipeline will result in natural gas flowing out of both ends of the break (downstream and upstream ends). The ALOHA computer program cannot model this scenario.

Consequently, Dresden (2023) has calculated an equivalent pipeline diameter to simulate the flow from both ends of the break of the TRL pipeline. The cross-sectional area of the pipeline with the equivalent diameter would be twice the cross-sectional area of the actual pipeline accounting for flow from both ends. As the inner diameter of the TRL pipeline is 19.25 in.

[48.9 cm] accounting for the pipeline wall thickness, the equivalent inner pipeline diameter would be 27.23 in. [69.2 cm] (Dresden, 2023). Resulting overpressure, thermal flux, and methane concentration are estimated using the flow of a pipeline having diameter equal to the equivalent diameter. The staff finds the approach acceptable and is consistent with other assessments, for example, Baker (2005).

Dresden (2023) has assumed that the TRL pipeline is connected to an infinite source of natural gas. In this analysis, the flow will continue for 60 min, as dictated by the ALOHA code. This means neither the automatic isolation valves at both ends of the pipeline, nor the Alliance Gas Control would isolate the pipeline for 60 minutes in case of a rupture. This is a conservative assumption, as given in Table 3, Gas Line Inputs, the TRL pipeline will be fitted with NPS 20 isolation valves (Dresden, 2023). Although these valves are designed to isolate a pipeline in approximately 20 s, a base case isolation valve close time of 10 min and a conservative case of 20 min are assumed by the Alliance Pipeline (Dresden, 2023, Attachment B).

Dresden (2023) has assumed the pipe length in the ALOHA model to be 1,000 ft [305 m] long connected to an infinite source. The staff finds this assumption of a short pipe length is conservative as assuming the actual pipe length of 2.85 mi [4.6 km] will significantly reduce the maximum release rate (by 57%) and the total amount released (by 69%) over the simulation period of 60 minutes because of higher pipe wall friction from a longer pipeline, as shown in the ALOHA runs.

4.2.3 Overpressure Estimates Dresden (2023) has estimated the overpressure at three selected locations of the Dresden Station from a hypothetical rupture of the TRL pipeline and subsequent release of the natural gas for 60 min using the ALOHA computer program.

9 The maximum overpressure experienced at the Meteorological Tower, the closest location to the TRL pipeline, is 0.374 psi [2.58 kPa]. The nearest plant structure would experience 0.11 psi

[0.76 kPa]. Overpressure at the Control Room is insignificant.

4.2.4 Staff Assessment The ALOHA computer program can model the release from a pipeline either (1) the pipeline is connected to an infinite source, or (2) the unbroken end of the pipeline is closed off. In Case (1),

the pipeline is never isolated either by automatic actuation of the isolation valves or by the Alliance Gas Control after the pipeline rupture. In Case (2), either the valves have automatically isolated, or the Gas Control have actuated isolation as soon as the pipeline ruptures. Based on Table 3 of Dresden (2023), neither scenarios model the reality. The isolation valves are designed to isolate the pipeline in approximately 20 s. Based on the recommendation given in FM-42 (Factory Mutual, 2008), it is assumed that the isolation valves would close within 10 min of rupture. A conservative case of 20 min of flow before the isolation valves actuate is also assumed. As discussed before, Dresden (2023) has assumed that the valves do not close for 60 min (connected to an infinite source), as controlled by the ALOHA program.

Based on the above discussion, the likely scenario is that the flow will continue at the maximum release rate after the rupture for 10 min (pipeline attached to an infinite source), followed by closure of the isolation valves. Thereafter, the release will be from a closed off pipeline. Only the natural gas within the pipeline would be released out the broken end till the inventory is exhausted. The staff has used the actual length of the pipeline, 15,048 ft or 2.85 mi [4.6 km].

Using the ALOHA program, the staff has estimated the total natural gas released as the sum of flow at the maximum flow rate assuming the unbroken end is connected to an infinite source for 10 min and, thereafter, the pipeline is connected to a closed off source releasing till the inventory within the pipeline is over. The staff has also repeated the analysis assuming 20-min of release before the unbroken end is assumed to be closed off. In both estimation, the staff has assumed the pipeline has an equivalent inner diameter of 27.23 in. [69.2 cm]. The total natural gas released is estimated to be 1,250,000 lb [566,990 kg] for 10-min and 2,500,000 lb

[1,133,981 kg] for 20-min of release after the rupture before closure of the isolation valves.

Using the results of the ALOHA computer program, the staff finds that the entire content of 207,160 lb [93,966 kg] in the 2.85 mi [4.6 km] long pipeline at the moment of closure of the isolation valves will be released within 14 min. Therefore, the total gas released after rupture would be the sum of first 10 min of flow with the infinite source and next 14 min of flow with source close off condition, which is 1,457,160 lb [660,957 kg]. For a 20-min initial release, the total release would be 2,707,160 lb [1,227,947 kg].

The staff has used the BST equations, as given in Tang and Baker (1999) with additional input for flame speed given in Pierorazio, et al. (2005), to estimate the overpressure at the three locations of the Dresden Station. Results are given in Table 1 and have been compared with those from Dresden (2023). In estimating the overpressure, the staff has assumed that the entire quantity of the released natural gas remains within the flammable range (5 to 15%

concentration for methane), conservative assumption ignoring any dilution in the atmosphere. In reality, the vapor cloud will disperse in air under the prevailing atmospheric conditions and the methane concentration will continue to dilute, as shown in the ALOHA runs (Dresden, 2023).

10 There will be a rich region near the rupture, a lean region at the clouds leading edges, and a flammable region in between.

Table 1. Estimated Overpressure at the Dresden Station Location Distance from Pipeline (ft/m)

(from Figures 2-4 of Dresden, 2023)

Estimated Overpressure (psi/kPa) 10 min before valve closure 20 min before valve closure Dresden (2023)

Meteorological Tower 1,259/384 0.077/0.531 0.083/0.572 0.374/2.579 Nearest Plant Structure 3,366/1,026 0.029/0.200 0.035/0.241 0.11/0.76 Control Room 3,701/1,128 0.026/0.179 0.032/0.221 Insignificant In summary, based on the preceding discussion, the staff finds that the overpressure estimated by Dresden (2023) for all locations of interest at the Dresden Station is significantly less than 1 psi [6.9 kPa]. As per RG 1.91 (NRC, 2021), 1 psi [6.9 kPa] is the threshold overpressure for damage to the safety-related structures. Therefore, the staff finds that a hypothetical rupture of the proposed TRL pipeline will not pose a credible hazard to the safety-related structures at the Dresden Station. All safety-related structures within the Dresden Station including the non-safety-related Meteorological Tower can withstand the overpressure levels without sustaining any damage.

4.3 Thermal Hazard A rupture of the proposed TRL natural gas transmission pipeline can develop a jet/crater fire near the rupture location and radiates heat flux if the released gas is ignited immediately after the rupture. Dresden (2023) has estimated the heat flux from the jet/crater fire and assessed the potential consequences to the Dresden Station. The staff has reviewed the analysis and assessment of associated consequences. The review is summarized below.

4.3.1 Thermal Flux Estimates Using the ALOHA computer program, Dresden (2023) has estimated the extent of the heat flux experienced by the surrounding area of the pipeline rupture location. Results are given for a 60-s exposure of at three levels of heat flux with associated effects (Dresden, 2023). The heat flux is estimated at three heat intensity levels:

634 Btu/hr-ft2 [2.0 kW/m2]: causes pain to human 1,586 Btu/hr-ft2 [5.0 kW/m2]: second degree burns 3,172 Btu/hr-ft2 [10.0 kW/m2]: potentially lethal.

The spatial extent of the heat flux at these intensity levels are given in Figure 7, Jet Fire, of Dresden (2023). The heat flux at locations beyond 1,932 ft [589 m] is less than 634 Btu/hr-ft2

[2 kW/m2]. Therefore, the nearest plant structure and the Control Room would not even experience a hear flux that could cause pain to human after 60 s of exposure. The Meteorological Tower at 1,250 ft [381 m] from the pipeline would experience a heat flux of

11 1,567 Btu/hr-ft2 [4.94 kW/m2]. Therefore, Dresden (2023) has concluded that this level of heat load would not compromise the integrity of the tower structure.

4.3.2 Staff Assessment The staff has reviewed the estimated heat flux levels at the Dresden Station from the ALOHA analysis. The staff finds that the PIR for this pipeline is 508 ft [155 m]. No significant impact on people and property would be expected beyond this PIR distance. All the three selected locations at the Dresden ISFSI are at least 1,250 ft [381 m] or further away. The estimated heat flux at the Control Room would be significantly less than 634 Btu/hr-ft2 [2 kW/m2]. Therefore, the staff accepts that the heat flux from a potential rupture of the TRL pipeline will have negligible effects on the Control Room operators. In addition, the nearest plant structures would receive a heat flux of less than 634 Btu/hr-ft2 [2 kW/m2]. A minimum radiant heat flux of at least 4,000 Btu/hr-m2 [12.5 kW/m2] would be needed to ignite wood with a flame or melt plastic tubing (Sluder et al., 2022). Steel structures are expected to deform after 30 min of exposure of 7,930 Btu/hr-ft2 [25 kW/m2]. The Meteorological Tower is expected to experience a heat flux of approximately 1,586 Btu/hr-ft2 [5 kW/m2]. Therefore, based on the preceding discussion, the staff agrees that the heat flux from the jet/crater fire following the rupture of the proposed TRL pipeline will not pose a credible hazard to the Dresden Station.

4.4 Methane Concentration at the Control Room Location Dresden (2023) has assessed the potential exposure of the Control Room operators to the released methane from the hypothetical rupture of the proposed TRL pipeline. Methane concentration at the outdoor and indoor locations of the Control Room of the Dresden Station have been estimated using the ALOHA computer program and compared with the limits specified by the Protective Action Criteria (PAC) for uncontrolled release of methane. At the PAC-1 level (65,000 ppm for methane), the general population when exposed for more than an hour could experience discomfort, irritation effects; however, these effects are not disabling.

They are temporary and reverse upon cessation of the exposure. At the PAC-2 level (230,000 ppm for methane), the general population exposed more than an hour could experience irreversible or other serious, long-lasting adverse effects. At the concentration level PAC-3 (400,000 ppm for methane) and above, the general population could experience life-threatening adverse health effects even death after an hour of exposure.

4.4.1 Methane Concentration Estimates Figure 6 of Dresden (2023) shows the methane concentration at the Control Room. The estimated maximum concentration at the outdoor is 7,120 ppm. The indoor concentration is 4,800 ppm, estimated assuming that the entire volume of the Control Room would be exchanged with intake air 1.48 times in an hour, as stated in Table 2 of Dresden (2023). The outdoor methane concentration is 21,400 ppm at the Meteorological Tower, closest to the pipeline. PAC-1 methane concentration extends up to 525 ft [160 m] in the downwind direction from the rupture point, as shown in Figure 5 of Dresden (2023). The PAC-2 and PAC-3 concentrations extend 330 ft [100 m] and 300 ft [91 m], respectively.

4.4.2 Staff Assessment The staff has reviewed both the estimated methane concentrations at the outdoor and indoor of the Control Room. The concentration at the outdoor location reaches 7,120 ppm after

12 approximately 20 min of rupture and remains constant during the remaining duration of the event, assumed 1 hr in the ALOHA analysis, as shown in Figure 6, Concentration at Control Room, of Dresden (2023). The concentration inside the Control Room gradually increases to 4,800 ppm within 1 hr of the rupture. This gradual increase is due to constant mixing of the outdoor air with higher methane content with the air inside the Control Room by the ventilation system. As the outdoor methane concentration remains constant, the indoor methane concentration gradually increases.

The staff also finds that the maximum outdoor methane concentration of 7,120 ppm is low, approximately 11% of the PAC-1 limit. At the PAC-1 concentration level, the general population could experience temporary discomfort which would be reversible after the end of the event.

Therefore, based on the preceding discussion, the staff finds that both the outdoor and indoor methane concentrations at the Control Room would be acceptable and are not expected to cause any serious, long-lasting health effects.

5.0

SUMMARY

AND CONCLUSIONS In summary, the annual frequency of rupture of the proposed TRL pipeline is estimated to be small. The estimated rupture frequency is also quite conservative as the model parameters used to estimate the annual rupture frequency are conservative. Nevertheless, the Alliance Pipeline L.P. estimated the resulting thermal flux and overpressure in the highly unlikely event that a jet/crater fire or an explosion of the released natural gas cloud would take place after a hypothetical rupture of the proposed pipeline. The maximum overpressure generated by either a 10-min or a 20-min natural gas release from both ends of the ruptured TRL pipeline under different confinement conditions would be significantly lower than 1 psi [6.9 kPa], the threshold for damage to safety-related structures (NRC, 2022). The safety-related structures at the Dresden Station are designed to withstand 3 psi [20 kPa] pressure drop from tornado wind. The thermal flux received by these safety-related structures from a jet/crater fire resulted from the rupture of the TRL pipeline would be significantly lower than the threshold radiant heat flux of 5,000 Btu/hr-ft2 [15.8 KW/m2] used as the basis to determine the PIR distance, as per 49 CFR Part 912.903. In addition, the duration of the heat flux would be short. Therefore, based on its review of the submitted analyses, the staff finds that the frequency to generate a jet/crater fire or an overpressure event from hypothetical rupture of the proposed TRL pipeline at its closest point to the Dresden Station is very low. These rare events are not expected to generate either an overpressure or a thermal flux that would be damaging to the safety-related structures. The overpressure and/or the thermal flux are expected to be well within the design of the safety-related structures of the Dresden Station.

In addition, the estimated overpressure and thermal flux at the Dresden Station from a hypothetical rupture of the TRL pipeline are smaller than the threshold overpressure and thermal flux to cause damage to the safety-related structures. Consequently, the exposure distance, as explained in Appendix A, Exposure Distance Calculation, of the Regulatory Guide (RG) 1.91 (NRC, 2021) would be zero for both overpressure and thermal flux. Therefore, the annual frequency of rupture of the proposed TRL pipeline would have negligible influences on the safety-related structures at the Dresden Station from the consequences of resulting overpressure and thermal flux. These structures are adequately designed for the potential overpressure and thermal hazards from rupture of the TRL pipeline.

13 The staff finds the analyses presented have several conservative assumptions:

1. The assumption that the potential rupture of the TRL pipeline will result in a guillotine break is conservative. This is the worst-case scenario of pipeline rupture. A more likely scenario would be a leak of the pipeline. The failure rate reported in J. House Environmental (2004) shows this.

2. The assumption that the rupture of the pipeline would occur at the worst location-the point closest to the Dresden Station, is conservative. The rupture can take place anywhere of the 2.85 mi (4.6 km) long pipeline.

3. The assumption that both ends of the ruptured pipeline will continuously release the natural gas at the constant rate till the valves have isolated the pipeline is conservative. The actual release rate at the downstream end of the pipeline will be rapidly reduced.

4. Although the TRL pipeline will be fitted with valves designed to isolate the pipeline in approximately 20 s, the analysis assumed continuous flow of natural gas through the ruptured section of the pipeline for 60 min.

5. Dresden (2023) has assumed that the TRL pipeline is only 1,000 ft [333 m] long, instead of 2.85 mi [4.6 km]. As a result, the frictional drag experienced by the natural gas from this short pipeline is much smaller than actual resulting in the estimated total quantity of release significantly more than actual.

6. An extremely rare combination of atmospheric conditions (Pasquill Stability Class F) has been assumed which only happens 0.39% of the time at the Dresden Power Station. This rare atmospheric conditions make the vapor cloud dispersion difficult resulting in higher quantity of natural gas within the flammable range. The estimate overpressure and heat load from ignition of this vapor cloud would be conservative.

6.0 REFERENCES

49 CFR Part 192.903. What Definitions Apply to the Subpart?

Atkinson, G., J. Hall, and A. McGillivray. 2017. RR1113. Review of Vapour Cloud Explosion Incidents. Health and Safety Executive, Derbyshire SK17 9JN. United Kingdom.

Baker, M. Jr., Inc. 2005. Potential Impact Radius Formulae for Flammable Gases Other Than Natural Gas Subject to 49 CFR 192: Final Report. TTO Number 13, Integrity Management Program. Research and Special Programs Administration, Office of Pipeline Safety. Department of Transportation. June.

Center for Chemical Process Safety (CCPS). 2014. Guidelines for Determining the Probability of Ignition of a Released Flammable Mass. American Society of Chemical Engineers, Hoboken, New Jersey, Wiley.

Dresden. 2023. Dresden Buried Pipe Blast and Gas Leak Analysis. Revision 001. Analysis No.

DRE22-0002. September 7.

14 Dresden. 2009. UFSAR Chapter 2.0 Site Characteristics. Dresden Power Station.

Factory Mutual Insurance Company. 2008. Guidelines for Evaluating the Effects of Vapor Cloud Explosions Using a TNT Equivalency Method. FM Global Property Loss Prevention Data Sheets 7-42.

NRC. 2004. Fire Dynamics Tools (FDTs): Quantitative Fire Hazard Analysis Methods for the U.S. Nuclear Regulatory Commission Fire Protection Inspection Program, Final Report.

NUREG-1805. Washington, DC. ADAMS Accession Number ML043290075.

Mercx, W.P.M. and A.C. van den Berg. 2005. Vapour Cloud Explosions. Chapter 5 of Methods for the Calculation of Physical Effects Due to Release of Hazardous Materials (Liquids and Gases). CPR 14E. Editors: C.J.H. van der Bosch and R.A.P.M. Weterings. The Director-General for Social Affairs and Employment. The Netherlands.

NRC. 2020. Report to the U.S. Nuclear Regulatory Commission Expert Evaluation Team on Concerns Pertaining to Gas Transmission Lines Near the Indian Point Nuclear Power Plant.

Washington, DC. ADAMS Accession Number ML20100F635.

Sluder, S., S. Rose, B. Oland, D. Sulfredge, R. Lee, and A. Coates. 2022. Updated Safety Review and Assessment of Natural Gas Transmission Pipelines Adjacent to the Indian Point Site. ORNL/SR-2022/2558. Oak Ridge National Laboratory. Oak Ridge, Tennessee.

Pierorazio, A.J., J.K. Thomas, Q.A. Baker, and D.E. Ketchum. 2005. An Update to the Baker-Strehlow-Tang Vapor Cloud Explosion Prediction Methodology Flame Speed Table. Process Safety Progress. Vol. 24, No. 1.

Tang, M.J. and Q.A. Baker. 1999. A New Set of Blast Curves from Vapor Cloud Explosion.

Process Safety Progress. Vol. 18, No. 3.

U.S. Nuclear Regulatory Commission (NRC). 2021. Evaluations of Explosions Postulated To Occur on Transportation Routes Near Nuclear Power Plants. Regulatory Guide (RG) 1.91.

Washington, DC. ADAMS Accession Number ML2260A242.

U.S. Nuclear Regulatory Commission (NRC). 2007. Design-Basis Tornado and Tornado Missiles for Nuclear Power Plants. Regulatory Guide (RG) 1.76. Washington, DC. ADAMS Accession Number ML070360253.